AI workloads fuel shift to private cloud

ITPro reports that a growing number of enterprises are turning back to private cloud, not just to manage costs, but to gain tighter control over security, compliance, and AI workloads, according to new research from GTT Communications.

While cost predictability and vendor lock-in remain concerns, GTT’s survey of U.S. and European enterprise leaders indicates that over half of AI workloads are now hosted in private cloud or on-premise environments, driven largely by data privacy and compliance pressures. Spending on private cloud infrastructure exceeding $10 million is expected to jump from 43% to 54% this year, outpacing public cloud investment growth nearly twofold. GTT VP Bastien Aerni noted that companies are adopting hybrid and multi-cloud strategies to balance performance and AI demands. However, 43% of enterprises reported difficulties repatriating applications and data from public cloud, underscoring the complexity of supporting AI at scale without reengineering network and security architecture. “Many still underestimate the complexity involved,” Aerni warned.